Stefano Esposito

I am a final-year PhD student at the University of Tübingen, part of the Autonomous Vision Group led by Prof. Andreas Geiger, and I work in close collaboration with Meta Reality Labs through ELLIS.

Research: I work on 3D and 4D reconstruction from images and video. My current research bridges learned generative priors with per-scene optimization: feed-forward and pixel-aligned generative 3D reconstruction (LaRa, GenIA), and data-driven efficient 3D Gaussian Splatting optimizers (Learn2Splat). Previously, I worked on neural fields and layered mesh representations for efficient, high-quality novel view synthesis (Volumetric Surfaces).

Bio: Before my PhD, I obtained a bachelor's degree in computer science and a master's degree in Artificial Intelligence and Robotics at Sapienza University of Rome, and worked as a research associate at Hochschule Bonn-Rhein-Sieg in Germany. Outside research, I'm a tech enthusiast, dog dad, and outdoors lover.

For any inquiries, feel free to reach out!

Publications

Highlighted papers are the most relevant to my current research.

GenIA: Generative Reconstruction with Test-Time Input Alignment
Stefano Esposito, Naama Pearl, Polina Karpikova, Samuel Rota Bulò, Lorenzo Porzi, Peter Kontschieder, Andreas Geiger, Jonathon Luiten
arXiv.org, 2026

Project Page Paper Code

@InProceedings{esposito2026genia, 
	author = {Stefano Esposito and Naama Pearl and Polina Karpikova and Samuel Rota Bulò and Lorenzo Porzi and Peter Kontschieder and Andreas Geiger and Jonathon Luiten}, 
	booktitle = {arXiv.org}, 
	title = {GenIA: Generative Reconstruction with Test-Time Input Alignment}, 
	year = {2026}, 
}

Learn2Splat: Extending the Horizon of Learned 3DGS Optimization
Naama Pearl^*, Stefano Esposito^*, Haofei Xu, Amit Peleg, Patricia Gschoßmann, Lorenzo Porzi, Peter Kontschieder, Gerard Pons-Moll, Andreas Geiger
arXiv.org, 2026 (* equal contribution)

Project Page Paper Code

@InProceedings{pearl2026learn2splat, 
	author = {Naama Pearl and Stefano Esposito and Haofei Xu and Amit Peleg and Patricia Gschoßmann and Lorenzo Porzi and Peter Kontschieder and Gerard Pons-Moll and Andreas Geiger}, 
	booktitle = {arXiv.org}, 
	title = {Learn2Splat: Extending the Horizon of Learned 3DGS Optimization}, 
	year = {2026}, 
}

ConeGS: Error-Guided Densification Using Pixel Cones for Improved Reconstruction with Fewer Primitives
Bartłomiej Baranowski, Stefano Esposito, Patricia Gschoßmann, Anpei Chen, Andreas Geiger
Proc. of the International Conf. on 3D Vision (3DV), 2026

Project Page Paper Code

@InProceedings{baranowski2026conegs, 
	author = {Bartłomiej Baranowski and Stefano Esposito and Patricia Gschoßmann and Anpei Chen and Andreas Geiger}, 
	booktitle = {Proc. of the International Conf. on 3D Vision (3DV)}, 
	title = {ConeGS: Error-Guided Densification Using Pixel Cones for Improved Reconstruction with Fewer Primitives}, 
	year = {2026}, 
}

Volumetric Surfaces: Representing Fuzzy Geometries with Layered Meshes
Stefano Esposito, Anpei Chen, Christian Reiser, Samuel Rota Bulò, Lorenzo Porzi, Katja Schwarz, Christian Richardt, Michael Zollhoefer, Peter Kontschieder, Andreas Geiger
Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2025

Project Page Paper Code

@InProceedings{Esposito2025VolSurfs, 
	author = {Stefano Esposito and Anpei Chen and Christian Reiser and Samuel Rota Bulò and Lorenzo Porzi and Katja Schwarz and Christian Richardt and Michael Zollhoefer and Peter Kontschieder and Andreas Geiger}, 
	booktitle = {Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR)}, 
	title = {Volumetric Surfaces: Representing Fuzzy Geometries with Layered Meshes}, 
	year = {2025}, 
}

LaRa: Efficient Large-Baseline Radiance Fields
Anpei Chen, Haofei Xu, Stefano Esposito, Siyu Tang, Andreas Geiger
Proc. of the European Conf. on Computer Vision (ECCV), 2024

Project Page Paper Code

@InProceedings{Chen2024ECCV, 
	author = {Anpei Chen and Haofei Xu and Stefano Esposito and Siyu Tang and Andreas Geiger}, 
	booktitle = {Proc. of the European Conf. on Computer Vision (ECCV)}, 
	title = {LaRa: Efficient Large-Baseline Radiance Fields}, 
	year = {2024}, 
}

Efficient Generation of Multimodal Fluid Simulation Data
Daniele Baieri, Donato Crisostomi, Stefano Esposito, Filippo Maggioli, Emanuele Rodolà
Smart Tools and Applications in Graphics (STAG), 2024

Project Page Paper

@InProceedings{Baieri2024STAG, 
	author = {Daniele Baieri and Donato Crisostomi and Stefano Esposito and Filippo Maggioli and Emanuele Rodolà}, 
	booktitle = {Smart Tools and Applications in Graphics (STAG)}, 
	title = {Efficient Generation of Multimodal Fluid Simulation Data}, 
	year = {2024}, 
}

Fluid Dynamics Network: Topology-Agnostic 4D Reconstruction via Fluid Dynamics Priors
Daniele Baieri, Stefano Esposito, Filippo Maggioli, Emanuele Rodolà
arXiv.org, 2023

Paper

@InProceedings{Baieri2023ARXIV, 
	author = {Daniele Baieri and Stefano Esposito and Filippo Maggioli and Emanuele Rodolà}, 
	booktitle = {arXiv.org}, 
	title = {Fluid Dynamics Network: Topology-Agnostic 4D Reconstruction via Fluid Dynamics Priors}, 
	year = {2023}, 
}

KiloNeuS: A Versatile Neural Implicit Surface Representation for Real-Time Rendering
Stefano Esposito, Daniele Baieri, Stefan Zellmann, Emanuele Rodolà, André Hinkenjann
arXiv.org, 2022

Paper

@InProceedings{Esposito2022ARXIV, 
	author = {Stefano Esposito and Daniele Baieri and Stefan Zellmann and Emanuele Rodolà and André Hinkenjann}, 
	booktitle = {arXiv.org}, 
	title = {KiloNeuS: A Versatile Neural Implicit Surface Representation for Real-Time Rendering}, 
	year = {2022}, 
}

Stefano Esposito

Publications

Software